Back to results
US 10,762,087 B2Granted
Database search
AssigneeEMC IP Holding Company LLC
InventorsYu Cao, Baoyao Zhou, Qiyan Chen, Bo Chen, Yuxi Sun
Life of this patent
Estimating the term…
Abstract
The present disclosure provides a method and apparatus for database search, by grouping data entries based on join conditions between the data entries as set in search conditions; and executing the search based on the grouping of the data entries, and can efficiently and effectively resolve the issues that are common to the existing MapReduce query processing systems, thereby being particularly suitable for big dataset analytics in a large cluster system.
Claims20
A method for database search, comprising:
- grouping data entries based on join conditions between the data entries as set in search conditions;
- implementing the search conditions using an extended-SQL query, wherein the extended SQL query is enabled to include nested queries;
- creating a query graph based on the grouping of data entries and the extended-SQL query, wherein a node in the query graph represents a query variable in the extended-SQL query and an edge between nodes in the query graph represents a joint condition between query variables;
- creating, based on the query graph, a query header derived from the extended-SQL query;
- analyzing the data entries to determine whether multiple MapReduces are enabled to be used to search;
- executing the search based on the grouping of the data entries and the query graph;
- and
- transforming the search conditions by removing patterns.
5 more elements
An apparatus for database search, comprising:
- a module configured to group data entries based on join conditions between the data entries as set in search conditions, wherein the module implements the search conditions using an extended-SQL query, wherein the extended SQL query is enabled to include nested queries, wherein the module is enabled to analyze the data entries to determine whether multiple MapReduces are enabled to be used to search;
- wherein the module is enabled to create a query graph based on the grouping of data entries and the extended-SQL query, wherein a node in the query graph represents a query variable in the extended-SQL query and an edge between nodes in the query graph represents a join condition between query variables and create, based on the query graph, a query header derived from the extended-SQL query;
- execute the search based on the grouping of the data entries and the query graph;
- and
- a pattern removing module configured to remove patterns from the search conditions.
2 more elements
Dependencies are inferred from the claim text.